4 easy tips for writing better AI prompts | Kunalsinh Kathia | TEDxSaffrony Institute of Technology

Key Concepts

• Prompting: The act of providing instructions to an AI; framed as an ancient human skill rather than a technical one.
• RISE Framework: A structured methodology for prompting (Role, Instructions, Steps, Expectations).
• AI Literacy: The ability to communicate effectively with machines, considered the "new literacy" of the modern era.
• Systematic Instruction: The necessity of breaking down complex tasks into logical sequences for AI processing.

---

1. The Core Problem: The "10% Instruction" Fallacy

The speaker argues that users often provide vague, minimal instructions (10%) while expecting perfect, high-quality results (100%).

• The Mirror Analogy: AI lacks human intuition, common sense, and the ability to "read between the lines." It acts as a mirror; if the output is poor, it reflects the lack of clarity in the user's input.
• Statistical Context: Research indicates that approximately 44.6% of AI failures are directly attributable to poor prompting rather than a lack of machine intelligence.

2. Prompting as a Human Skill

Contrary to the belief that prompting requires coding or mathematical expertise, the speaker defines it as an extension of human communication.

• Historical Context: Humans have been "prompting" each other for centuries (e.g., a king instructing a minister or a teacher assigning work).
• The Shift: While humans can infer intent from tone and facial expressions, AI is literal. Therefore, users must transition from "deep thinking" (internal) to "explicit articulation" (external) to bridge the gap between human intent and machine execution.
• Jensen Huang’s Perspective: The CEO of Nvidia notes that "the new programming language is the human," meaning the most important code one can write is a clear, well-structured thought.

3. The RISE Framework

To move from being a passive "user" to an active "director," the speaker introduces the RISE framework:

• R (Role): Assign a persona to the AI (e.g., "You are an expert physics teacher"). This forces the AI to adopt a specific perspective and knowledge base.
• I (Instructions): Be direct and specific. Avoid vague requests like "tell me about science" in favor of precise tasks like "explain the concept of gravity."
• S (Steps): Break the task into a logical sequence. This prevents the AI from rushing and ensures it follows a path, reducing the likelihood of errors.
• E (Expectations): Define the format and constraints of the output (e.g., a 5-minute script, a story, or a bulleted list). This eliminates the AI's need to "guess" the desired format.

4. Practical Application: The Ohm’s Law Case Study

The speaker demonstrates the framework by contrasting a "lazy prompt" with a "RISE prompt" for teaching Ohm’s Law ($V=IR$) to 40 sleepy students:

• Lazy Prompt: "Give me an activity to teach Ohm's law." (Result: Boring, theoretical, and impractical).
• RISE Prompt:
  • Role: Expert physics teacher who makes science fun.
  • Instructions: Create a physical classroom activity for $V=IR$.
  • Steps: Use a water analogy; assign roles (voltage/resistance) to students; provide a 5-minute summary script.
  • Expectations: No lab equipment required; classroom-ready.
• Outcome: A highly engaging, actionable, and instrument-free lesson plan.

5. Synthesis and Conclusion

The speaker concludes that prompting is not merely a tool-based skill but a cognitive discipline.

• Actionable Insight: By learning to prompt, individuals are training their own brains to think with greater clarity, structure, and intention.
• Final Takeaway: The future does not belong to the smartest machines, but to the humans who possess the literacy to direct them effectively. One must stop being a passive user and start acting as a director of the technology.